Non-insulin dependent diabetes mellitus (NIDDM) afflicts 4-5 million Americans every year. NIDDM is treated predominately with insulin. However, insulin is not convenient to use in that it must be injected 2-4 times per day and must be stored properly to prevent loss of efficacy. Other drugs used to treat NIDDM include troglitazone (Rezulin.TM.), a PPAR.gamma. agonist, Glucophage.TM. and sulfonylureas. Unfortunately, there are safety concerns related to the use of these drugs. The identification of safe, effective, orally available drugs for the treatment of NIDDM would greatly enhance the quality of life of patients who suffer from this disease. However, studies to find such molecules have been hampered by a lack of reproducible human in vitro cell systems.
Approximately 20-25% of Americans are obese and therefore at increased risk for developing NIDDM, hypertension, and cardiovascular disease. The causes of NIDDM and obesity are often related to defects or problems with adipose tissue. Adipocytes play a critical role in lipid storage and metabolism. Adipocytes also act as endocrine cells to influence physiological parameters such as insulin sensitivity and body weight (Flier, et al., Cell, (1995) 80:15-18)). For example, the ob gene encodes leptin, an adipocyte-secreted endocrine factor (Zhang, et al., Nature (1994) 372:425-432). Leptin has been shown to reduce body weight and blood glucose in obese, diabetic rodents (Pelleymounter, et al., Science, (1995) 269:540-543).
Several adipocyte-specific enzymes and receptors have been shown to be important targets for anti-obesity and anti-diabetic drug discovery. For example, agonists of the .beta.3 adrenergic receptor, which is found predominantly in the adipose tissue in man (Amer, et al., New England Journal of Medicine, (1995) 333:382-383), have anti-obesity and anti-diabetic properties in rodents and are currently in phase II/III trials in man. The thiazolidinedione class of compounds (TZDs), including troglitazone and ciglitazone, has been shown to improve insulin sensitivity and thereby reduce hyperglycemia and hyperlipidemia conditions in rodents and in humans (Saltiel, et al., Diabetes, (1996) 45:1661-1669; Sreenan, et al., American Journal Physiol, (1996) 271:E742-E747; Nolan, et al., New England Journal of Medicine, (1994) 331:1188-1193. Troglitazone (Rezulin.TM.) is approved for use in the U.S. and Japan. Many TZDs, including troglitazone and ciglitazone, are potent activators of Peroxisome Proliferator Activated Receptor gamma (PPAR.gamma.), a member of the nuclear receptor family of transcription factors (Tontonoz, et al., Cell, (1994) 79:1147-1156; Lehmann, et al., Journal of Biological Chemistry, (1995) 270:12953-12955). PPAR.gamma. is a key regulator of adipocyte differentiation and is most abundant in adipose tissue.
Animal adipocyte studies have been facilitated by the availability of a number of immortalized preadipocyte cell lines such as the 3T3-L1 mouse fibroblast line (Green, et al., Cell, (1974) 1:113-116), which upon proper treatment, will differentiate into adipocytes. These cells have many of the same properties as isolated primary adipocytes. However, recent published reports show that human adipose tissue and adipocytes exhibit significant differences from rodent cells with respect to factors affecting insulin resistance. For example, TNF.alpha. appears to be regulated differently in human adipose tissue than in rodent adipose tissue (Hotamisligil, et al., Journal of Clin Invest, (1995) 95:2409-2415). Consequently, studies on human adipocytes and adipose metabolism have been hampered by the lack of a preadipocyte cell culture that can be reproducibly induced to differentiate into adipocytes at high efficiency.
Current protocols for differentiating isolated human preadipocytes result in differentiation frequencies of 5-80%. In these systems, the preadipocyte cell component in human adipose tissue (the so-called "stromal vascular fraction" or SVF) can be isolated using collagenase treatment (Rodbell, Journal of Biol Chem., (1967) 242:5744-5750; Rodbell, et al., Meth Enzymol, (1974) 31:103-14). The isolated human preadipocytes can then be driven to differentiate into adipocytes by a variety of chemical treatments. For example, Hauner's laboratory (Hauner, et al., Journal Clin Invest., (1989) 34:1663-1670) has shown that human preadipocytes can be induced to differentiate in serum-free medium containing 0.2 nM triiodothyronine, 0.5 .mu.M insulin and 0.1 .mu.M glucocorticoid (cortisol, dexamethasone or aldosterone). Under these conditions, differentiation of 5-70% of the preadipocytes was achieved. The percentage of differentiated cells was related to the age of the subject from which the cells were obtained. These investigators claim they can achieve anywhere from 5-70% complete differentiation within 20 days as determined by a variety of biochemical markers. Similarly, O'Rahilly's laboratory (Digby, et al., Diabetes, (1998) 5:138-141) have shown that when the above serum-free medium is supplemented with a TZD such as BRL 49653, differentiation of 20% of omental preadipocytes and 50-80% of subcutaneous and perirenal preadipocytes is achieved.
A method of differentiating human preadipocytes to adipocytes at higher frequency in a shorter period of time and with greater consistency would aid in the study of obesity and diabetes. The present invention provides methods and compositions for the consistent differentiation of 90-95% of human preadipocytes. Several methodologies are provided for different and selective applications.